Tests based on join count statistics and on Monte-Carlo simulations were developed to characterize the aggregation and propagation patterns of esca-symptomatic vines.
Geo-statistical methods
The prediction of sparsely measured Available Water Capacity using densely measured resistivity data have been performed in two steps, using a combination of several methods such as median polish kriging.
Spatio-Temporal modelling
Hierarchical models have been developed to describe the dynamic of spatial disease structure, bayesian inferences were done by INLA.
Two-step centered autologistic model.
Internship works
Time series
To test and fit the periodicity of pine processionary moth outbreaks using periodogram ks test and a sinusoide autoregressive model respectively.